Coupling

The "soft link" coupling method (Wene 1996), involves keeping the models’ full structure and complexity, exchanging a chosen set of variables and solving the models iteratively until convergence is reached on a given criterion. It has the advantage of allowing the use of detailed and complex models, which we deem important for an analysis of the impact of climate and energy policies on the electricity sector and the entire economy. It permits for the representation of the electricity sector’s

25 Developed by the Energy Economics Group at the Laboratory for Energy Systems Analysis (LEA), Paul Scherrer Institute (PSI), Switzerland.

95/174

interaction with the other sectors of the economy (Schäfer and Jacoby 2005), and allows for different types of policies to be adequately modeled: market instruments in the CGE model and technology standards in the bottom-up model.

Coupling through a soft link prioritizes the strengths of each model: The electricity generation mix and costs from the bottom-up model are given priority over the electricity production function of the CGE model. The latter effectively becomes a Leontief function which is parametrized with informa-tion from the latest bottom-up run. On the other hand, the endogenous electricity demand reacinforma-tion of the top-down model is given precedence over the initial demand assumption for the electricity supply model. Additionally, the variations of factor and intermediate input price variations due to general equilibrium effects modify the investment costs, and operation and maintenance costs of the bottom-up model.

Figure 49 depicts the exchange of information between the two models. Electricity generation costs and their components as well as export revenues and import costs are extracted from the CROSSTEM-CH model and translated for the CGE model into a) the wholesale electricity price and b) input shares for factors and commodities to the electricity generation cost function. The sectoral²⁶ electricity demand quantities simulated by the GENESwIS model are then sent back to become inputs to the CROSSTEM-CH model. To account for changes in the economy, factor and intermediate input prices from GENESwIS are used to modify the investment costs and operation and maintenance costs of the different technologies in the bottom-up model. This sequence is iterated upon until the vector of quantities of total electricity demanded each year converges.

Figure 49: Information exchange between the two component models

26 GENESwIS simulates yearly electricity demands that are distributed to each of the 288 time slices of the CROSSTEM-CH with the help of sectoral load curves.

96/174

The modeling framework has been set up such that the link between CROSSTEM-CH’s generation costs and GENESWIS’ wholesale prices for electricity can be modeled in different ways:

• Average cost plus pricing: The wholesale price is set at CROSSTEM-CH’s average cost level plus a markup.

• Marginal cost pricing: CROSSTEM-CH’s marginal cost is the shadow price of the commodity balance and represents the increase in total system cost due to an additional unit of demand (Loulou et al. 2005). It reflects all constraints and costs (incl. investment cost) and can therefore be seen as a long-term marginal cost, or marginal cost including scarcity rents for capacity. As the CGE model does not disaggregate the year into 288 time slices, the marginal cost is aggregated to an annual demand-weighted marginal cost.

To help with the convergence of the models, which is hampered by the stepwise behavior of the bottom-up supply curve, we introduce a supply elasticity in the Electricity Transport and Distribution sector of the CGE model. For this purpose, we insert a fixed resource at the top of the Electricity Transport and Distribution’s nest. The elasticity of substitution between the fixed resource and the rest of the inputs is calculated²⁷ such that, given the share of the fixed resource, the supply elasticity of the sector equals a selected value²⁸ (see Rutherford 1998). This method was inspired by the work of Lanz and Rausch 2011 who introduce a demand elasticity to parameterize the bottom-up demand.

They show that the choice of demand elasticity does not affect the results but that a good approxi-mation of the top-down demand response reduces the number of iterations needed for converg-ence.

Despite the introduction of a supply elasticity in the Electricity Transport and Distribution sector of the CGE model, it is frequent for the models to lock up into an oscillation between two marginal costs. The electricity demand oscillates between two values and does not converge. To avoid this problem, we introduce a dampening of the demand response in the coupler: Instead of the last electricity demand vector, we send a Gauss-Seidel combination of the CGE electricity demands of the previous iterations (equation 3.1.9) to the bottom-up model:

(3.1.9) where α ∈ [0, 1] represents the length of the step towards the demand of the last iteration.

WP 3 Scenarios

We simulate a baseline and a policy scenario for two different types of electricity markets in Switzer-land: a regulated market, and progressive liberalization to a fully liberalized market (see Table 22).

27 , with σ the elasticity of substitution between the fixed resource and the rest of the inputs, η the supply elasticity and θR the share of the fixed resource.

28 The value of supply elasticity can be set such that it helps with convergence (approximating the bottom-up supply elasticity), as it has no impact on the results.

97/174

Table 22: Scenarios matrix

The baseline (BAU) scenarios are based on the weiter wie bisher (i.e. "more of the same") scenario of the Energy Perspectives 2050 (Prognos 2012). They include current policies such as an Emissions Trading Scheme, a CO2 tax on gas and heating fuels for the non-ETS sectors and households, and a subsidy program for the energy refurbishment ofbuildings. For each pricing scenario, the GENESwIS model is calibrated such that the electricity demands and CO2 emissions follow the paths projected by Prognos 2012.

The TAX scenarios represent more stringent climate and energy policies. A tax is levied on electricity at a rate of 10% in 2020, increasing linearly to 50% in 2050. The Emissions Trading Scheme stays identical as in the BAU scenario, but the CO2 tax on gas and heating fuels is increased linearly from current level (60 CHF/t) to 200 CHF/t in 2050. A CO2 tax on transport fuels is introduced at 50 CHF/t in 2035, reaching 200 CHF/t in 2050.

Under regulation (scenario REG), firms are usually allowed to cover their costs and make an appro-priate profit. We assume accordingly that electricity is priced at average cost plus a small markup.

We assume in the liberalized market scenario (LIB) that the electricity market will be entirely libera-lized from 2025 onwards and that the price will then follow the long-term marginal cost of the bottom-up model. From 2010 to 2025, the market is in transition and prices reflect an increasing importance of marginal cost pricing. Profit is calculated such that the price of wholesale electricity is pushed from the average cost given by the CROSSTEM-CH model (AC) to the assumed market price (Pm).

(3.1.10) We analyze the policy scenarios for the two market regulation assumptions TAX_LIB and TAX_REG as deviations from the respective baseline scenarios BAU_LIB and BAU_REG. It is uncommon in a CGE setting to have two different baselines. Actually, the central targeted baseline parameters, namely electricity demands and CO2 emissions per fuels, are the same in both of our baselines. However, it was necessary to recalibrate the model framework under the different coupling mechanisms to match the targeted baseline parameters, as electricity prices are defined in a different manner.